From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugzilla@gcc.gnu.org>
Received: by sourceware.org (Postfix, from userid 48)
 id 76F93386EC49; Sat,  6 Mar 2021 22:50:19 +0000 (GMT)
DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 76F93386EC49
From: "pinskia at gcc dot gnu.org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug target/99434] std::bit_cast generates more instructions than
 __builtin_bit_cast and memcpy with -march=native
Date: Sat, 06 Mar 2021 22:50:19 +0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: changed
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: target
X-Bugzilla-Version: 11.0
X-Bugzilla-Keywords: missed-optimization, ra
X-Bugzilla-Severity: enhancement
X-Bugzilla-Who: pinskia at gcc dot gnu.org
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Resolution: 
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags: 
X-Bugzilla-Changed-Fields: 
Message-ID: <bug-99434-4-wZqlP7x4KR@http.gcc.gnu.org/bugzilla/>
In-Reply-To: <bug-99434-4@http.gcc.gnu.org/bugzilla/>
References: <bug-99434-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-BeenThere: gcc-bugs@gcc.gnu.org
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Gcc-bugs mailing list <gcc-bugs.gcc.gnu.org>
List-Unsubscribe: <https://gcc.gnu.org/mailman/options/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=unsubscribe>
List-Archive: <https://gcc.gnu.org/pipermail/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-request@gcc.gnu.org?subject=help>
List-Subscribe: <https://gcc.gnu.org/mailman/listinfo/gcc-bugs>,
 <mailto:gcc-bugs-request@gcc.gnu.org?subject=subscribe>
X-List-Received-Date: Sat, 06 Mar 2021 22:50:19 -0000

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=3D99434

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
(In reply to cqwrteur from comment #2)
> (In reply to Andrew Pinski from comment #1)
> > This is just a register allocation issue dealing with mulx and TImode.
> >=20
> > If mulq was used instead (that is without -march=3Dnative), all of the
> > functions are done correctly.
>=20
> I do not think so. I think GCC generally did things like this wrong. I ha=
ve
> even found out how to produce different wrong results deterministically.
>=20
> For example like this
> https://godbolt.org/z/PbobYG
>=20
> Any time it deals with things like >>32 or >>64, it produces a slower res=
ult.
> This even compiles without -march=3Dnative.

This is still a register allocation issue, this time dealing with DImode on
32bit. GCC has a known issue with register allocation when dealing with val=
ues
stored into two registers.  See PR 21150, PR 43644, PR 50339, etc.=